| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Fri, 13 Dec 2013 20:10:05 +0100
From: Jan Kara <jack@...e.cz>
To: Toralf Förster <toralf.foerster@....de>
Cc: Richard Weinberger <richard.weinberger@...il.com>,
Jan Kara <jack@...e.cz>,
"linux-mm@...ck.org" <linux-mm@...ck.org>,
Linux Kernel <linux-kernel@...r.kernel.org>,
UML devel <user-mode-linux-devel@...ts.sourceforge.net>
Subject: Re: [uml-devel] why does index in truncate_inode_pages_range() grows
so much ?
On Fri 13-12-13 14:01:34, Toralf Förster wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA256
>
> On 12/13/2013 11:51 AM, Richard Weinberger wrote:
> > On Fri, Dec 13, 2013 at 10:52 AM, Toralf Förster <toralf.foerster@....de> wrote:
> >> -----BEGIN PGP SIGNED MESSAGE-----
> >> Hash: SHA256
> >>
> >> On 12/11/2013 09:26 PM, Jan Kara wrote:
> >>> Thanks! So this works more or less as expected - trinity issued a
> >>> read at absurdly high offset so we created pagecache page a that
> >>> offset and tried to read data into it. That failed. We left the
> >>> page in the pagecache where it was for reclaim to reclaim it when
> >>> free pages are needed. Everything works as designed except we could
> >>> possibly argue that it's not the most efficient way to use
> >>> pages...
> >>>
> >>> Patch 'vfs: fix a bug when we do some dio reads with append dio
> >>> writes' (http://www.spinics.net/lists/linux-fsdevel/msg70899.html)
> >>> should actually change the situation and we won't unnecessarily
> >>> cache these pages.
> >>>
> >> confirmed - applied to latest git tree of Linus I helps.
> >
> > Good to know! :-)
> >
>
> OTOH - there's seems to be more places for an improvement - now trinity
> often runs hours w/o problem (before it runs within a rather short time
> into such issues).
>
> But today I got another case (I did not patched the source files except
> the mentioned patch by Jan Kara) where the trinity call cycles since 2
> hours w/o any progress. But fortunately the system is still responsive,
> ssh works and I can shutdown that virtual machine smoothly, furthermore
> all local and remote file systems can be unounted cleanly - so that patch
> is a big improvement).
OK, that is indeed strange. Can you step through find_get_pages() in gdb
to see where exactly are we looping? From the traces below the argument
values look somewhat unreliable so it's not clear to me what is exactly
happening.
Honza
> tfoerste@n22 ~/devel/github/bingo $ date; sudo gdb /home/tfoerste/devel/linux/linux 9776 -n -batch -ex 'bt'
> Fri Dec 13 13:54:33 CET 2013
> find_get_pages (mapping=0x45182810, start=0, nr_pages=14, pages=0x0) at mm/filemap.c:885
> 885 }
> #0 find_get_pages (mapping=0x45182810, start=0, nr_pages=14, pages=0x0) at mm/filemap.c:885
> #1 0x080d669a in pagevec_lookup (pvec=0x40607d40, mapping=0x0, start=0, nr_pages=0) at mm/swap.c:937
> #2 0x080d6a9a in truncate_inode_pages_range (mapping=0x45182810, lstart=0, lend=-1) at mm/truncate.c:241
> #3 0x080d6e3f in truncate_inode_pages (mapping=0x0, lstart=0) at mm/truncate.c:358
> #4 0x0818c2d2 in ext4_evict_inode (inode=0x45182758) at fs/ext4/inode.c:228
> #5 0x0811b5ff in evict (inode=0x45182758) at fs/inode.c:549
> #6 0x0811c0ed in iput_final (inode=<optimized out>) at fs/inode.c:1419
> #7 iput (inode=0x45182758) at fs/inode.c:1437
> #8 0x08112056 in do_unlinkat (dfd=5, pathname=0x8065d84 <register_lines+276> "") at fs/namei.c:3718
> #9 0x081121c5 in SYSC_unlinkat (flag=<optimized out>, pathname=<optimized out>, dfd=<optimized out>) at fs/namei.c:3754
> #10 SyS_unlinkat (dfd=5, pathname=134634884, flag=0) at fs/namei.c:3746
> #11 0x08062a94 in handle_syscall (r=0x473b59cc) at arch/um/kernel/skas/syscall.c:35
> #12 0x080750f5 in handle_trap (local_using_sysemu=<optimized out>, regs=<optimized out>, pid=<optimized out>) at arch/um/os-Linux/skas/process.c:198
> #13 userspace (regs=0x473b59cc) at arch/um/os-Linux/skas/process.c:431
> #14 0x0805f750 in fork_handler () at arch/um/kernel/process.c:149
> #15 0x5a5a5a5a in ?? ()
>
>
> tfoerste@n22 ~/devel/github/bingo $ date; sudo gdb /home/tfoerste/devel/linux/linux 9776 -n -batch -ex 'bt'
> Fri Dec 13 13:54:47 CET 2013
> radix_tree_next_chunk (root=0x3f, iter=0x40607cdc, flags=6) at lib/radix-tree.c:773
> 773 index &= ~((RADIX_TREE_MAP_SIZE << shift) - 1);
> #0 radix_tree_next_chunk (root=0x3f, iter=0x40607cdc, flags=6) at lib/radix-tree.c:773
> #1 0x080cc88e in find_get_pages (mapping=0x45182810, start=0, nr_pages=14, pages=0x6) at mm/filemap.c:844
> #2 0x080d669a in pagevec_lookup (pvec=0x40607d40, mapping=0x3f, start=63, nr_pages=63) at mm/swap.c:937
> #3 0x080d6a9a in truncate_inode_pages_range (mapping=0x45182810, lstart=0, lend=-1) at mm/truncate.c:241
> #4 0x080d6e3f in truncate_inode_pages (mapping=0x3f, lstart=25769803839) at mm/truncate.c:358
> #5 0x0818c2d2 in ext4_evict_inode (inode=0x45182758) at fs/ext4/inode.c:228
> #6 0x0811b5ff in evict (inode=0x45182758) at fs/inode.c:549
> #7 0x0811c0ed in iput_final (inode=<optimized out>) at fs/inode.c:1419
> #8 iput (inode=0x45182758) at fs/inode.c:1437
> #9 0x08112056 in do_unlinkat (dfd=5, pathname=0x8065d84 <register_lines+276> "") at fs/namei.c:3718
> #10 0x081121c5 in SYSC_unlinkat (flag=<optimized out>, pathname=<optimized out>, dfd=<optimized out>) at fs/namei.c:3754
> #11 SyS_unlinkat (dfd=5, pathname=134634884, flag=0) at fs/namei.c:3746
> #12 0x08062a94 in handle_syscall (r=0x473b59cc) at arch/um/kernel/skas/syscall.c:35
> #13 0x080750f5 in handle_trap (local_using_sysemu=<optimized out>, regs=<optimized out>, pid=<optimized out>) at arch/um/os-Linux/skas/process.c:198
> #14 userspace (regs=0x473b59cc) at arch/um/os-Linux/skas/process.c:431
> #15 0x0805f750 in fork_handler () at arch/um/kernel/process.c:149
> #16 0x5a5a5a5a in ?? ()
>
>
> tfoerste@n22 ~/devel/github/bingo $ date; sudo gdb /home/tfoerste/devel/linux/linux 9776 -n -batch -ex 'bt'
> Fri Dec 13 13:57:16 CET 2013
> radix_tree_next_chunk (root=0x10, iter=0x40607cdc, flags=6) at lib/radix-tree.c:769
> 769 while (++offset < RADIX_TREE_MAP_SIZE) {
> #0 radix_tree_next_chunk (root=0x10, iter=0x40607cdc, flags=6) at lib/radix-tree.c:769
> #1 0x080cc88e in find_get_pages (mapping=0x45182810, start=0, nr_pages=14, pages=0x6) at mm/filemap.c:844
> #2 0x080d669a in pagevec_lookup (pvec=0x40607d40, mapping=0x10, start=16, nr_pages=16) at mm/swap.c:937
> #3 0x080d6a9a in truncate_inode_pages_range (mapping=0x45182810, lstart=0, lend=-1) at mm/truncate.c:241
> #4 0x080d6e3f in truncate_inode_pages (mapping=0x10, lstart=25769803792) at mm/truncate.c:358
> #5 0x0818c2d2 in ext4_evict_inode (inode=0x45182758) at fs/ext4/inode.c:228
> #6 0x0811b5ff in evict (inode=0x45182758) at fs/inode.c:549
> #7 0x0811c0ed in iput_final (inode=<optimized out>) at fs/inode.c:1419
> #8 iput (inode=0x45182758) at fs/inode.c:1437
> #9 0x08112056 in do_unlinkat (dfd=5, pathname=0x8065d84 <register_lines+276> "") at fs/namei.c:3718
> #10 0x081121c5 in SYSC_unlinkat (flag=<optimized out>, pathname=<optimized out>, dfd=<optimized out>) at fs/namei.c:3754
> #11 SyS_unlinkat (dfd=5, pathname=134634884, flag=0) at fs/namei.c:3746
> #12 0x08062a94 in handle_syscall (r=0x473b59cc) at arch/um/kernel/skas/syscall.c:35
> #13 0x080750f5 in handle_trap (local_using_sysemu=<optimized out>, regs=<optimized out>, pid=<optimized out>) at arch/um/os-Linux/skas/process.c:198
> #14 userspace (regs=0x473b59cc) at arch/um/os-Linux/skas/process.c:431
> #15 0x0805f750 in fork_handler () at arch/um/kernel/process.c:149
> #16 0x5a5a5a5a in ?? ()
>
> - --
> MfG/Sincerely
> Toralf Förster
> pgp finger print:1A37 6F99 4A9D 026F 13E2 4DCF C4EA CDDE 0076 E94E
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.22 (GNU/Linux)
> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>
> iF4EAREIAAYFAlKrBS0ACgkQxOrN3gB26U4bRgD/U/5sIELFBZUTeEgfM9eJBnxh
> PhdMMBTTJHoB3v9z70YA/iEZzD9L30vVSWqYrybOWNPYwDR1i67F41nUemmPczqu
> =u/iT
> -----END PGP SIGNATURE-----
--
Jan Kara <jack@...e.cz>
SUSE Labs, CR
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists